The aim of the GENCODE consortium is to annotate all evidence-based gene features in the human genome at a high accuracy, including protein-coding loci with alternatively splices variants, non-coding loci and pseudogenes. With this proposal we aim to extend GENCODE to the mouse genome and use the comparison of corresponding human and mouse loci to improve both sets of annotation. Despite the tremendous progress of current GENCODE production project, and the current outstanding quality of at least the protein-coding gene set, a complete annotation of all human genes is far from complete. For example, it has recently become clear that the number of non-coding RNA genes is far greater than previously supposed. It is also recognized that there are still substantial numbers of alternative transcripts still to be discovered from transcriptomics studies of additional cell types. Our first aim is therefore to continue to improve the coverage and accuracy of the GENCODE human gene set. Our second aim is to apply to the mouse genome the same annotation approaches as we have applied to human to generate the human GENCODE gene set. To achieve both goals we will integrate computational approaches, expert manual annotation and targeted experimental approaches as we have done for human. We will also use comparative approaches to use the resulting mouse annotation to inform and improve the human GENCODE gene set. A comprehensive knowledge of the location and structure of genes in the human genome is central to our understanding of human biology and the mechanisms of disease. Similarly for mouse, a comprehensive high quality gene set will aid in the design of experiments and the interpretation of the effects of gene knockouts and resulting phenotypes. Also, since mouse is used as a model of human, knowledge of its genes and their relationship to human genes will help inform human gene function. The outputs of regular releases of GENCODE gene sets will therefore be of benefit to the entire community of human and mouse researchers.